If people are to make wise decisions in our heavily technological world, is it essential that they learn how to code?
For author and artist James Bridle, that is analogous to asking whether it is essential that people be taught plumbing skills.
Of course we want and need people who know how to connect water taps, how to find and fix leaks. But,
“learning to plumb a sink is not enough to understand the complex interactions between water tables, political geography, ageing infrastructure, and social policy that define, shape and produce actual life support systems in society.” (Except where otherwise noted, all quotes in this article are from New Dark Age by James Bridle, Verso Books 2018)
Likewise, we need people who can view our technological society as a system – a complex, adaptive and emergent system – which remains heavily influenced by certain motives and interests while also spawning new developments that are beyond any one group’s control.
Bridle’s 2018 book New Dark Age takes deep dives into seemingly divergent subjects including the origins of contemporary weather forecasting, mass surveillance, airline reservation systems, and Youtube autoplay lists for toddlers. Each of these excursions is so engrossing that it is sometimes difficult to hold his central thesis in mind, and yet he weaves all the threads into a cohesive tapestry.
Bridle wants us to be aware of the strengths of what he terms “computational thinking” – but also its critical limitations. And he wants us to look at the implications of the internet as a system, not only of power lines and routers and servers and cables, but also of people, from the spies who tap into network nodes to monitor our communications, to the business analysts who devise ways to “monetize” our clicks, to the Facebook groups who share videos backing up their favoured theories.
Wiring of the SEAC computer, which was built in 1950 for the U.S. National Bureau of Standards. It was used until 1964, for purposes including meteorology, city traffic simulations, and the wave function of the helium atom. Image from Wikimedia Commons.
From today’s weather, predict tomorrow’s
Decades before a practical electronic computer existed, pioneering meteorologist Lewis Fry Richardson1 thought up what would become a “killer app” for computers.
Given current weather data – temperature, barometric pressure, wind speed – for a wide but evenly spaced matrix of locations, Richardson reasoned that it should be possible to calculate how each cell’s conditions would interact with the conditions in adjacent cells, describe new weather patterns that would arise, and therefore predict tomorrow’s weather for each and all of those locations.
That method became the foundation of contemporary weather forecasting, which has improved by leaps and bounds in our lifetimes. But in 1916, when Richardson first tried to test his ideas they were practically useless. The method involved so many calculations that Richardson worked for weeks, then months, then years to work out a ‘prediction’ from a single day’s weather data.
But by the end of World War II, the US military had developed early electronic computers which could begin to make Richardson’s theory a useful one. To military strategists, of course, the ability to predict weather could provide a great advantage in war. Knowing when a particular attack would be helped or hindered by the weather would be a great boon to generals. Even more tantalizingly, if it were possible to clearly understand and predict the weather, it might then also be possible to control the weather, inflicting a deluge or a sandstorm, for example, on vulnerable enemy forces.
John von Neumann, a mathematician, Manhattan Project physicist and a major figure in the development of computers, summed it up.
“In what could be taken as the founding statement of computational thought, [von Neumann] wrote: ‘All stable processes we shall predict. All unstable processes we shall control.’”
Computational thinking, then, relied on the input of data about present conditions, and further data on how such conditions have been correlated in the past, in order to predict future conditions.
But because many aspects of our world are connected in one system – an adaptive and emergent system – this system spawns new trends which behave in new ways, not predictable simply from the patterns of the past. In other words, in the Anthropocene age our system is not wholly computable. We need to understand, Bridle writes, that
“technology’s increasing inability to predict the future – whether that’s the fluctuating markets of digital stock exchanges, the outcomes and applications of scientific research, or the accelerating instability of the global climate – stems directly from these misapprehensions about the neutrality and comprehensibility of computation.”
Take the case of climate studies and meteorology. The technological apparatus to collect all the data, crunch the numbers, and run the models is part of a huge industrial infrastructure that is itself changing the climate (with the internet itself contributing an ever-more significant share of greenhouse gas emissions). As a result the world’s weather is ever more turbulent, producing so-called ‘100 year storms’ every few years. We can make highly educated guesses about critical climatic tipping points, but we are unable to say for sure when these events will occur or how they will interact.
Age-old traditional knowledge of ways to deal with this week’s or this year’s weather is becoming less reliable. Scientists, too, should acknowledge the limits of computational thinking for their work:
“In a 2016 editorial for the New York Times, computational meteorologist and past president of the American Meteorological Society William B. Gail cited a number of patterns that humanity has studied for centuries, but that are disrupted by climate change: long-term weather trends, fish spawning and migration, plant pollination, monsoon and tide cycles, the occurrence of ‘extreme’ weather events. For most of recorded history, these cycles have been broadly predictable, and we have built up vast reserves of knowledge that we can tap into in order to better sustain our ever more entangled civilisation.”
The implications are stark: “Gail foresees a time in which our grandchildren might conceivably know less about the world in which they live than we do today, with correspondingly catastrophic events for complex societies.”
World map of submarine communication cables, 2015. Cable data by Greg Mahlknecht, world map by Openstreetmap contributors. Accessed through Wikimedia Commons.
Lines of power
In many ways, Bridle says, we can be mislead by the current view of the internet as a “cloud”. Contrary to our metaphor, he writes, “The cloud is not weightless; it is not amorphous, or even invisible, if you know where to look for it.” To be clear,
“It is a physical infrastructure consisting of phone lines, fibre optics, satellites, cables on the ocean floor, and vast warehouses filled with computers, which consume huge amounts of water and energy and reside within national and legal jurisdictions. The cloud is a new kind of industry, and a hungry one.”
We have already referred to the rapidly growing electricity requirements of the internet, with its inevitable impact on the world’s climate. When we hear about “cloud computing”, Bridle also wants us to bear in mind the ways in which this “cloud” both reflects and reinforces military, political and economic power relationships:
“The cloud shapes itself to geographies of power and influence, and it serves to reinforce them. The cloud is a power relationship, and most people are not on top of it.”
It is no accident, he says, that maps of internet traffic trace pathways of colonial power that are hundreds of years old. And we shouldn’t be surprised that the US military-intelligence complex, which gave birth to internet protocols, have also installed wiretapping equipment and personnel at junctions where trans-oceanic cables come ashore in the US, allowing them to scoop up far more communications data than they can effectively monitor.2
These power relationships come into play in determining not only what is visible in our web applications, but what is hidden. Bridle is a keen plane-spotter, and he marvels at flight-tracking websites which show, in real time, the movements of thousands of commercial aircraft around the world. “The view of these flight trackers, like that of Google Earth and other satellite image services, is deeply seductive,” he says, but wait:
“This God’s-eye view is illusory, as it also serves to block out and erase other private and state activities, from the private jets of oligarchs and politicians to covert surveillance flights and military manoeuvres. For everything that is shown, something is hidden.”
Aviation comes up frequently in the book, as its military and commercial importance is reflected in the outsize role aviation has played in the development of computing and communications infrastructure. Aviation provides compelling examples of the unintended, emergent consequences of this technology.
High anthropoclouds in the sky of Barcelona, 2010, accessed through Wikimedia Commons. The clouds created by aircraft have an outsize impact on climate change. And climate change, Bridle writes, contributes to the increasingly vexing problem of “clear air turbulence” which threatens aircraft but cannot be reliably predicted.
On the last day of October, just a few months after New Dark Age was published, I found myself at Gatwick International Airport near London. I wanted to walk to the nearby town of Crawley to pick up a cardboard packing box. Though the information clerks in the airport terminal told me there was no walking route to Crawley, I had already learned that there was in fact a multi-use cycling lane, and so I hunted around the delivery ramps and parking garage exits until I found my route.
It was a beautiful but noisy stroll, with a brook on one side, a high fence on the other, and the ear-splitting roar of jet engines rising over me every few minutes. Little did I know that in just over a month this strange setting would be a major crime scene, as the full force of the aeronautical/intelligence industry pulled out all stops to find the operators of unauthorized drones, while hundreds of thousands of passengers were stranded in the pre-Christmas rush.
Another month has passed and no perpetrators have been identified, leading some to wonder if the multiple drone sightings were all mistakes. But in any case, aviation experts have long agreed that it’s just a matter of time before “non-state actors” manage to use unmanned aerial vehicles to deadly effect. Wireless communications, robotics, and three-dimensional location systems are now so widely available and inexpensive, it is unrealistic to think that drones will always be controlled by or even tracked by military or police authorities.
The exponential advance of artificial stupidity
Bridle’s discussion of trends in artificial intelligence is at once one of the most intriguing and, to this layperson at least, one of the less satisfying sections of the book. Many of us have heard about a new programming approach, following which a computer program taught itself to play the game Go, and soon was able to beat the world’s best human players of this ancient and complex game.
Those of us who have had to deal with automated telephone-tree answering systems, as much as we may hate the experience, can recognize that voice-recognition and language processing systems have also gotten better. And Google Translate has improved by leaps and bounds in just a few years time.
Bridle’s discussion of the relevant programming approaches presupposes a basic familiarity with the concept of neural networks. Since he writes so clearly about so many other facets of computational thinking, I wish he had chosen to spell out the major approaches to artificial intelligence a bit more for those of us who do not have degrees in computer science.
When he discusses the facility of Youtube in promoting mindless videos, and the efficiency of social media in spreading conspiracy theories of every sort, his message is lucid and provocative.
Here the two-step dance between algorithms and human users of the web produces results that might be laughable if they weren’t chilling. Likewise, strange trends develop out of interplay between Google’s official “mission” – “to organize the world’s information” – and the business model by which it boosts its share price – selling ads.
The Children’s Youtube division of Google has been one of Bridle’s research interests, and those of us fortunate enough not to be acquainted with this realm of culture are likely to be shocked by what he finds.
You might ask what kind of idiot would name a video “Surprise Play Doh Eggs Peppa Pig Stamper Cars Pocoyo Minecraft Surfs Kinder Play Doh Sparkle Brilho”. A clever idiot, that’s who, an idiot who may or may not be human, but who knows how to make money. Bridle explains the motive:
“This unintelligible assemblage of brand names, characters and keywords points to the real audience for the descriptions: not the viewer, but the algorithms that decide who sees which videos.”
These videos are created to be seen by children too young to be reading titles. Youtube accommodates them – and parents happy to have their toddlers transfixed by a screen – by automatically assembling long reels of videos for autoplay. The videos simply need to earn their place in the playlists with titles that contain enough algorithm-matching words or phrases, and hold the toddler’s attention long enough for ads to be seen and the next video to begin.
The content factories that churn out videos by the millions, then, must keep pace with current trends while spending less on production than will be earned by the accompanying ads, which are typically sold on a “per thousand views” basis.
Is this a bit of a stretch from “organizing the world’s information”? Yes, but what’s more important, a corporation’s lofty mission statement, or its commercial raison d’être? (That is, to sell ads.)
When it comes to content aimed at adults the trends are just as troubling, as Bridle’s discussion of conspiracy theories makes clear.
According to the Diagnostic and Statistical Manual of Mental Disorders, he explains, “a belief is not a delusion when it is held by a person’s ‘culture or subculture’.”
But with today’s social media, it is easy to find people who share any particular belief, no matter how outlandish or ridiculous that belief might seem to others:
“Those that the psychiatric establishment would have classified as delusional can ‘cure’ themselves of their delusions by seeking out and joining an online community of like minds. Any opposition to this worldview can be dismissed as a cover-up of the truth of their experience ….”
This pattern, as it happens, reflects the profit-motive basis of social media corporations – people give a media website their attention for much longer when it spools videos or returns search results that confirm their biases and beliefs, and that means there are more ads viewed, more ad revenue earned.
If Google and other social media giants do a splendid job of “organizing the world’s information”, then, they are equally adept at organizing the world’s misinformation:
“The abundance of information and the plurality of worldviews now accessible to us through the internet are not producing a coherent consensus reality, but one riven by fundamentalist insistence on simplistic narratives, conspiracy theories, and post-factual politics. It is on this contradiction that the idea of a new dark age turns: an age in which the value we have placed upon knowledge is destroyed by the abundance of that profitable commodity, and in which we look about ourselves in search of new ways to understand the world.”
Our unknowable future
After reading to the last page of a book in which the author covers a dazzling array of topics so well and weaves them together so skillfully, it would be churlish to wish he had included more. I would hope, however, that Bridle or someone with an equal gift for systemic analysis will delve into two questions that naturally arise from this work.
Bridle notes that the energy demands of our computational network are growing rapidly, to the point that this network is a significant driver of climate change. But what might happen to the network if our energy supply becomes effectively scarce due to rapidly rising energy costs?3
Major sectors of the so-called Web 2.0 are founded in a particular business model: services are provided to the mass of users “free”, while advertisers and other data-buyers pay for our attention in order to sell us more products. What might happen to this dominant model of “free services”, if an economic crash means we can’t sustain consumption on anything close to the current scale?
I suspect Bridle would say that the answers to these questions, like so many others, do not compute. Though computation can be a great tool, it will not answer many of the most important questions.
In the morass of information/misinformation in which our network engulfs us, we might find many reasons for pessimism. But Bridle urges us to accept and even welcome the deep uncertainty which has always been a condition of our existence.
As misleading as the “cloud” may be as a picture of our computer network, Bridle suggests we can find value if we take a nod from the 14th-century Christian mystic classic “The Cloud of Unknowing.” Its anonymous author wrote, “On account of pride, knowledge may often deceive you …. Knowledge tends to breed conceit, but love builds.”
Or in Bridle’s 21st century phrasing,
“It is this cloud that we have sought to conquer with computation, but that is continually undone by the reality of what we are attempting. Cloudy thinking, the embrace of unknowing, might allow us to revert from computational thinking, and it is what the network itself urges upon us.”
Photo at top: anthropogenic clouds over paper mill UPM-Kymmene, Schongau, 2013. Accessed at Wikimedia Commons.
NOTES
1 For an excellent account of the centuries-long development of contemporary meteorology, including the important role of Lewis Fry Richardson, see Bill Streever’s 2016 book And Soon I Heard a Roaring Wind: A Natural History of Moving Air.
2 More precisely, though intelligence agents can often zero in on suspicious conversations after a crime has been committed or an insurgency launched, the trillions of bits of data are unreliable sources of prediction before the fact.
3 Kris de Decker has posed some intriguing possibilities in Low-Tech Magazine. See, for example, his 2015 article “How to Build a Low-tech Internet”.
